Qwen3.8-27B-GGUF is an open-weight model from Advantech. Its published files total 89.1 GB. It draws 889 downloads a month.
Advantech-EIOT GGUF artifacts, organized by quantization and role: - Q4KM/: Q4KM main models - IQ3KT/: IQ3KT main models Note: The IQ3KT model has performance issues and is for testing purposes only.
Model Card
Advantech-EIOT GGUF artifacts, organized by quantization and role: - Q4KM/: Q4KM main models - IQ3KT/: IQ3KT main models Note: The IQ3KT model has performance issues and is for testing purposes only. - mmproj/: multimodal projector files Exact duplicate local files were removed by SHA-256 before upload.
Excerpt from the card by Advantech.
Identity and Version
- Repository
- Advantech-EIOT/Qwen3.8-27B-GGUF
- Publisher
- Advantech
- Task
- Not stated by the source
- Modality
- Other
- Library
- Not stated by the source
- Parameters
- Not stated by the source
- Languages
- Not stated by the source
- Revision
- 1ecf9614f56c476f173bb5d6f246266423413ef9
- First published
- 2026-09-09
- Last updated
- 2026-10-07
Files and Weights
11 files, 89.1 GB in total. The weights are 9 files totalling 89.1 GB in gguf.
Every file
| File | Type | Size | SHA-256 |
|---|---|---|---|
| IQ3_KT/Qwen3.8-27B-IQ3_KT-count-balanced.gguf | Weights | 11.2 GB | 0c20635847a4 |
| IQ3_KT/Qwen3.8-27B-IQ3_KT-outQ6.gguf | Weights | 11.8 GB | 9b3acde550c4 |
| IQ3_KT/Qwen3.8-27B-IQ3_KT.gguf | Weights | 11.2 GB | 47ae3d577102 |
| Q4_K_M/Qwen3.8-27B-AD-Q4_K_M.gguf | Weights | 17.1 GB | 9f21564b0c96 |
| Q4_K_M/Qwen3.8-27B-AD-Q5_K_M-Q4_K_M.gguf | Weights | 18.5 GB | b6122f58711f |
| Q4_K_M/Qwen3.8-27B-Q4_K_M.gguf | Weights | 16.5 GB | 322e194ff797 |
| mmproj/AtomicChat/mmproj-Qwen3.8-27B-BF16.gguf | Weights | 931.1 MB | 1b0c1111e44b |
| mmproj/custom/mmproj-Qwen3.8-27B-BF16.gguf | Weights | 931.1 MB | bedc0a76e327 |
| mmproj/mmproj-Qwen3.8-27B-F16.gguf | Weights | 927.6 MB | 2f0a90f14032 |
| README.md | Documentation | 385 B | — |
| .gitattributes | Repository | 2.2 KB | — |
License and Download
- License
- Not stated by the source
- Access
- Open weights, no gate
- Download size
- 89.1 GB
Released by Advantech through its official repository on Hugging Face.
Memory Requirements
| Precision | Weights in memory |
|---|---|
| As published | 89.1 GB |
Weights only, from the published parameter count; the key-value cache and runtime add to this.